Soluble Ziegler-Natta type catalysts for the polymerization of olefins are well known in the art. Generally, these catalysts comprise a Group IV-B metal compound and a metal alkyl cocatalyst, particularly an aluminum alkyl cocatalyst. A subgenus of such catalysts is that wherein the Group 4 metal component comprises a bis(cyclopentadienyl) Group 4 metal compound (i.e.--a "metallocene"), particularly a titanium compound, in combination with an aluminum alkyl cocatalyst. While speculation remains concerning the actual structure of the active catalyst species of this subaenus of soluble Ziegler-Natta type olefin polymerization catalysts, it appears generally accepted that the structure of the catalytically active species is a Group 4 metal cation in the presence of a labile stabilizing anion. This is a theory advocated by Breslow and Newburg, and Long and Breslow, in their respective articles in J. Am. Chem. Soc., 1959, Vol. 81, pp. 81-86, and J. Am. Chem. Soc., 1960, Vol. 82, pp. 1953-957. As there indicated, studies have suggested that the catalytically active species is a titanium-alkyl complex or a species derived therefrom when a titanium compound; viz., bis(cyclopentadienyl)titanium dihalide, and an aluminum alkyl are used as a catalyst or catalyst precursor. The presence of ions, all being in equilibrium, when a titanium compound is used was also suggested by Dyachkovskii, Vysokomol. Soved., 1965, Vol. 7, pp. 114-115 and by Dyachkovskii, Shilova and Shilov, J. Polym. Sci., Part C, 1967, pp. 2333-2339. That the active catalyst species is a cation complex when a titanium compound is used, was further suggested by Eisch et al., J. Am. Chem. Soc., 1985, Vol. 107, pp. 7219-7221.
While the foregoing articles teach or suggest that the active catalyst species is an ion pair wherein the Group IV-B metal component is present as a cation, all of the articles teach the use of a cocatalyst comprising a Lewis acid either to form or to stabilize the active ionic catalyst species. The active catalyst is, apparently, formed through a Lewis acid-Lewis base reaction of two neutral components (the metallocene and the aluminum alkyl), leading to an equilibrium between a neutral catalytically inactive adduct and the active catalyst ion pair. As a result of this equilibrium, there is a competition for the anion which must be present to stabilize the active cation catalyst species. This equilibrium is, of course, reversible and such reversal deactivates the active catalyst species. Further, many, if not all, of the Lewis acids heretofore contemplated for use in soluble Ziegler-Natta type catalyst systems are chain transfer agents and, as a result, prevent effective control of the product polymer molecular weight and molecular weight distribution. Still further, the catalyst systems heretofore proposed do not generally facilitate incorporation of a significant amount of a plurality of different monomers or random distribution of such monomers when used in copolymerization processes, particularly .alpha.-olefin copolymerization processes.
The aforementioned metallocene catalyst systems are not highly active, nor are they generally active when the Group 4 metal is zirconium or hafnium. More recently, however, active Ziegler-Natta type catalysts have been found which are formed when bis(cyclopentadienyl) compounds of the Group 4 metals, including zirconium and hafnium, are combined with alumoxanes. As is well known, these systems, particularly those employing a zirconocene, offer several distinct advantages, including much higher activities than the aforementioned bis(cyclopentadienyl)titanium catalysts and the production of polymers with narrower molecular weight distributions than those from conventional Ziegler-Natta catalysts. Achiral bis(cyclopentadienyl)hafnium compounds, hafnocenes, used with alumoxane cocatalysts have offered few, if any, advantages when compared to analogous titanocenes or zirconocenes with respect to catalyst activity, polymer molecular weights, or extent or randomness of comonomer incorporation. Giannetti, Nicoletti, and Mazzochi, J. Polym. Sci., Polym. Chem., 1985, Vol. 23, pp. 2117-2133, claim that the ethylene polymerization rates of hafnocenes are five to ten times slower than those of similar zirconocenes while there is little difference between the two metallocenes in the molecular weight of the polyethylene formed from them. European Patent Application No. 200,351 A2 (1986) suggests that in the copolymerization of ethylene and propylene, there is little difference between titanocenes, zirconocenes and hafnocenes either in polymer molecular weights and molecular weight distributions or in ability to incorporate propylene randomly. Recently, however, Ewen et al. discloses in J. Am. Chem. Soc., 1987, Vol. 109, pp. 6544-6545, that chiral hafnocenes used with an alumoxane cocatalyst give isotactic polypropylene of higher molecular weight than that obtained from analogous chiral zirconocenes. In light of the deficiencies of the metallocene catalyst systems heretofore contemplated, a need still exists for an improved metallocene catalyst system which: (1) permits better control of polymer product's molecular weight and molecular weight distribution.; (2) is not subject to activation equilibrium, and (3) does not require the use of an undesirable excess of the cocatalyst. The need for a catalyst system which will facilitate the production of higher molecular weight polymeric products and facilitate incorporation of a larger amount of comonomer into a copolymer is also believed to be readily apparent.